The goal of this project is the development of a working, contractile, 3-dimensional computer model of the heart, its valves, and the nearby great vessels. Such a model will provide a new research tool that should take its place alongside of clinical studies, animal experiments, and research in physical models as a fourthmode of investigation of the heart. The proposed model is based on a generalization of the Navier-Stokes equations of a Viscous, incompressible fluid. In this generalization, the heart muscle, the valve leaflet and the walls of the great vessles are modeled by a collection of contractile or elastic factors that are immersed in the fluid, that move at the local fluid velocity, and that exert forces locally on the fluid. The fluid equations are solved on a fixed regular computational lattice and the moving fiber points are coupled to this lattice by a computational model of the Dirac Gamma-function. Once the model is operational, it will be used to follow the development of the cardiac flow pattern during embryonic and fetal life, to explore the role of mechanical forces during development, to investigate the hemodynamics of congenital heart disease, to elucidate the functional anatomy of the heart and its valves, to determine the influence of the pattern of activation of the myocardium on the mechanics of contraction, to study the mechanical consequences of infarction in different regions of the heart, to improve the design of prosthetic heart valves through computer testing, and to compare the behavior of natural and prosthetic heart valves under under stressful conditions such as exercise or disease. Because these studies will be conducted in a computer model, the investigator will have the advantage of complete control over each individual parameter of the computational experiment and complete access to all computed variables as output. Some different modes of output that will be available from these computer experiments are three-dimensional stereoscopic movies showing the motion of the blood, the valve leaflets, and the heart walls; two-dimensional movies of the same phenomena in selected cross-sectional planes through the heart; and graphs of computed pressures, flows, and chamber dimensions as functions of time. Collectively, these modes of output will give a far more detailed view of the beating heart and its flow pattern than is available experimentally. At the same time, for comparison with clinical or experimental data, the computer model can also generate simulated phonocardiograms, ultrasonic sector scans, angiograms, nuclear medicine scans, or computed tomography scans of the beating heart. In summary, the proposed cardiac simulator will be a versatile tool for investigating the normal and pathological function of the heart and for improving the design of prosthetic heart valves.